Notes from the machine room.

Raffael builds automation-first platforms and edge systems by day, and tries out ideas in language tooling, local-first apps, simulations, and games on the side. Writing about platform automation, edge infrastructure, applied security, and AI-assisted engineering. Languages and models are tools. They serve humans, not the other way around.

Read the writingExplore the map

Latest writing

All articles →

  1. Observability Is Not Logging

    A cost total says nothing about how a run went. This part turns each agent run into a trace: which turn escalated, which tool failed, where the money went.

    AgentsAI platformsObservabilityOpenTelemetryTypeScript

    15 min

  2. This Site Is Becoming a Map

    A blog shows what I wrote last. The new map on raskell.io shows what I keep coming back to, and it is the start of a public second brain.

    This siteKnowledge managementThe open webZola

    9 min

  3. The Bill Is Part of the Loop

    Every model response already carries its own bill in the usage field, and the loop throws it away. This post reads that field, prices it against a table, and turns cost into a value the loop can see per turn and act on: a metering decorator that needs no loop change, a running ledger, prompt-cache accounting that moves the number the most, and a budget that degrades before it stops. Part eight of The Agent Platform Handbook.

    AgentsAI platformsEconomicsObservabilityMetricsTypeScript

    10 min

  4. The Model Is a Dependency

    Every layer of the harness has existed to put the right tokens in front of one hardcoded model string. This post pulls that string out and makes the model a dependency: why you route a cheap model for easy turns and a strong one for hard turns, what a fallback looks like when a provider is down, and why 'models are commodities' is a claim about your architecture before it is a claim about the market. Part seven of The Agent Platform Handbook.

    AgentsAI platformsInferenceReliabilityEconomicsTypeScript

    14 min

  5. The Bottleneck Is Data, and Europe Is Sitting on It

    Frontier AI is running short on public text, not compute. Europe's edge is the high-quality data it never put online. Capturing it needs coordination.

    SovereigntyPolicyEconomics

    34 min

  6. Brain Emulation, Read as an Infrastructure Problem

    An MIT thesis argues brain emulation has crossed into engineering. Read as infrastructure, the numbers describe a workload your AI cluster was not built for.

    Distributed systemsInferenceGPUs

    12 min

  7. Memory Is a Write Path

    Retrieval reads context the project authored. Memory is the agent writing context for its own future sessions, and the write is the part that is dangerous. Why session memory comes free from the transcript, why cross-session memory needs a separate write surface, and how to let an agent remember without letting it corrupt its own ground truth. Part six of The Agent Platform Handbook.

    AgentsContext engineeringKnowledge managementTypeScript

    19 min

  8. The Proof You Only Have to Write Once

    Generation is cheap, so verification is the bottleneck — and we throw the verification away every time. What if a proof, once written, could be named, shared, and vouched for? A content-addressed commons for verified Haskell, and the move I left for after the semantic-artifact trilogy.

    Formal methodsProgramming languagesAI-assisted engineeringOpen sourceHaskell

    13 min

What I am working on

Projects

All projects →